As the deputy chief technology officer at the White House from 2015 to the end of the Obama administration, Ed Felten has faced some challenging issues. Issues like election security, government systems vulnerability, and consumer data protection. “I actually did a little bit of policy work surrounding blockchain and cryptocurrencies, as well,” says Felten.

Today, Felten teaches computer science at Princeton and serves as the cofounder of Offchain Labs, a company focused on creating private and scalable smart contracts. It just raised $3.7 million in funding round led by Pantera Capital. Offchain’s main product, Arbitrum, is (take a deep breath) a decentralized, blockchain-agnostic Layer 2 scalability solution for businesses. Like many other solutions that rely on mutual agreement between validators (aka the people responsible for making sure the contracts execute correctly), its validation solution is complicated and requires a deep knowledge of both intricate tech and bureaucratic-style processes. It’s the sort of byzantine expertise one might have after serving as chief technologist at the U.S. Federal Trade Commission and working with the U.S. Department of Justice Antitrust Division, as Felten has.

During his time at the White House, conversations about cryptocurrency largely revolved around regulation. “There were a bunch of discussions about how the government might be able to use blockchain and cryptocurrency to build and operate some of the systems that it uses,” Felten says. “I found a cautiously pro-cryptocurrency attitude among a lot of people in the government.”

There were even talks about the government issuing its own cryptocurrency, a sort of ultimate stablecoin minted by the United States. “I don’t want to give the impression that anybody was close to doing such a thing,” Felten says.

Felten has no way of knowing whether the government has continued these conversations about crypto, but he’s certainly kept talking about blockchain-based systems in his own time, now aiming to apply them to the private sector. Felten started Offchain Labs in August 2018 with a couple of his PhD students, creating Arbitrum to sell to businesses (specifically in the gaming, finance, and consortium blockchain sectors, but Felten won’t name any names). The product targets businesses that a) don’t want to operate on public blockchains, and b) wish to process transactions faster than they can on the likes of bitcoin and Ethereum.

The initial Arbitrum product, set for an alpha release next month and a commercial release by the end of this summer, will work only with Ethereum, speeding up processes that would have otherwise had to run at a speed of 15 transactions per second on that blockchain. Arbitrum will do this by posting just one transaction on the Ethereum blockchain for every full sequence of operations running on a connected off-chain smart contract. The goal is for Arbitrum to eventually interoperate with any Layer 1 blockchain, like bitcoin.

Get the BREAKERMAG newsletter, a weekly roundup of blockchain business and culture.

“What a lot of people are trying to do, and what we think we do better than others, is to move the vast majority of the activity needed to ensure correctness of smart contracts into an off-chain setting, while at the same time providing strong guarantees of correctness,” Felten says. “That’s really the challenge all of these systems face.”

Let’s parse out what a “guarantee of correctness” means. When someone at a business sets out to launch a set of contracts on Arbitrum, they have to choose a set of validators. Arbitrum offers what Felten calls an “any trust guarantee,” meaning that if at least one validator behaves honestly, the contract will execute correctly.

“Even if all the others are in an evil conspiracy, you’re still good,” he says. “That’s in contrast to other systems where either you need a majority to be honest, or you need majority agreement, or you sometimes even need two-thirds agreement—or other systems where if you don’t get agreement, you have to go back and put everything on chain.”

If, say, two validators on Arbitrum disagree, they end up engaging in a “sort of game,” as Felten describes it, “that’s refereed by a [separate] Ethereum contract.” Skipping over some of the technical details (which include narrowing the dispute down to the exact step in the Arbitrum contract it’s referring to), this “game” requires both validators, the challenger and the asserter, to back their claims with crypto deposits. The asserter then has to provide proof that their version of the contract instruction is correct. The Ethereum contract determines whether that proof is valid, and the person who’s wrong loses their deposit. “So you have incentive not to lie,” says Felten.

Only if “every validator of your contract is in an evil conspiracy,” says Felten, do you have a problem. There’s no real failsafe for this possibility, he explains, because implementors of Arbitrum get to choose their own validators, and be validators themselves. In a business environment, where everyone is ostensibly working toward the same goal, this should make sense.

The checks and balances of Arbitrum’s system could remind someone of the bureaucratic processes of Felten’s earlier career. Only Arbitrum, if it works as planned, will run considerably faster.